Meta Hires Apple's Top AI Executive; California Bill to Regulate AI Companies Introduced
Key Takeaways
- Meta hired Apple's top AI model executive with a multi-million dollar annual package.
- California bill introduced to regulate AI companies, mandating safety protocol releases and incident reporting.
- Panasonic's ConnectAI reduced business operation hours by 448,000 by 2024, utilization rate at 49.1%.
- Researchers developed RisingAttacK, a method to deceive AI models by subtly altering visual input.
- Coop Kobe uses AI for efficient delivery routes, reducing truck mileage by 10%.
Top Stories
Meta Hires Apple's Top AI Executive.
Meta hired Apple's top AI model executive, offering a package worth tens of millions of dollars a year, highlighting growing competition in AI talent acquisition. Meta has been aggressively recruiting top AI researchers from rivals like OpenAI, Anthropic, and Google, offering multi-million-dollar deals and high pay packages.
California Bill to Regulate AI Companies Introduced.
California State Senator Scott Wiener introduced a bill on July 9, 2025, to regulate AI companies, mandating public release of safety protocols and reporting of critical safety incidents. Companies potentially affected include OpenAI, Google, and Anthropic.
Panasonic's ConnectAI Reduces Business Operation Hours.
Panasonic Connect introduced its AI assistant service "ConnectAI" in February 2023, leveraging large language models from OpenAI, Google, and Anthropic, resulting in a reduction of 448,000 hours in business operations by 2024. The monthly utilization rate reached 49.1%, a 14.3 percentage point increase from the previous year, with 2.4 million uses, a 71% increase from the previous year.
Researchers Develop RisingAttacK to Deceive AI Models.
Researchers at North Carolina State University developed RisingAttacK, a method to subtly alter visual input and deceive AI models, manipulating AI computer vision systems without changing the image's appearance to human observers. RisingAttacK successfully targeted top AI models used in cars, cameras, and healthcare diagnostics.
Coop Kobe Uses AI for Efficient Delivery Routes.
On July 7, 2025, Coop Kobe (Kobe City) introduced AI to propose efficient delivery routes, reducing truck mileage by 10% and enabling deliveries with fewer personnel. The AI system, developed by OptiMind (Nagoya University startup), creates delivery routes in minutes by setting conditions, even without experienced personnel.
AI Advancements
Japan's AI Adoption Triples, Still Lags U.S. and China.
Japan's 2025 Information and Communication White Paper indicates that generative AI adoption in Japan reached 26.7% in 2024, a threefold increase from 2023, but still lags behind the U.S. (68.8%) and China (81.2%). Usage is highest among those in their 20s at 44.7%.
SMBC Group Launches AI Solutions Company in Singapore.
Sumitomo Mitsui Financial Group (SMBC Group) established a new AI solutions company in Singapore, jointly funded with Ahmed Mazhari, former Microsoft Asia President, who will serve as an AI advisor. The initiative is part of SMBC's 500 billion yen generative AI investment plan, targeting $150 million (approximately ¥200-22 billion) in sales within five years.
Google Launches AI Mode in Search in India.
On July 8, 2025, Google launched AI Mode in Search for users in India, powered by Gemini 2.5, providing AI-generated responses based on web information and allowing complex queries using text, voice, or images. AI Mode builds upon AI Overviews, serving over 1.5 billion monthly users, enhancing search with reasoning and multimodal functionality.
Nvidia Stock Nears $4 Trillion Market Cap.
Nvidia's stock rose to $160 on July 8, nearing a $4 trillion market cap, with analysts predicting a $175 target, driven by AI demand. Nvidia is approaching a $4 trillion market cap, fueled by AI growth and market confidence.
Foxconn Aims to Become AI Software Company by 2030.
Foxconn, a Taiwanese electronics component manufacturer, aims to transform into an AI software company by 2030, establishing a new subsidiary, Think Robotic, to reduce dependence on the consumer electronics industry and mitigate risks associated with geopolitical tensions.
Dentsu Japan Launches AI Center.
On July 7, 2025, Dentsu launched the 'dentsu Japan AI Center' with about 1,000 AI specialists from five companies within the Dentsu Japan Group, aiming to drive AI transformation and enhance business efficiency by integrating technology, data, talent, and knowledge related to AI.
Capgemini to Acquire WNS for $3.3 Billion.
Capgemini is acquiring WNS for $3.3 billion, offering $76.50 per share in cash, which is a 28% premium over WNS's 90-day average, to expand AI capabilities, including generative and agentic AI. The deal, approved by both boards, is expected to close by year-end.
Starlink Aviation Provides Fast In-Flight Wi-Fi.
Starlink Aviation provides fast in-flight Wi-Fi, nearly twice as fast as other options, with median download speeds reaching 152 Mbps and upload speeds 24 Mbps, with 44ms latency. Starlink's advantage stems from its fleet of 7,000 low-Earth orbit satellites.
Google's Gemini Now Analyzes Videos.
Google's Gemini now analyzes videos, summarizing content and detecting issues, allowing users to upload videos and ask questions for detailed explanations. The tool provides timestamped summaries and trivia, enabling users to quickly find specific information and create quizzes.
OpenAI Hires Engineers from Rivals to Accelerate AI Development.
OpenAI hired four high-profile engineers from rival companies, including David Lau, former VP of software engineering at Tesla, to accelerate the 'Stargate' project, focusing on infrastructure and research teams to accelerate AI development. This move occurs amid intense AI talent competition, following Meta's recent recruitment of seven OpenAI employees.
Verizon Provides Satellite Devices to Texas DPS.
On July 7, 2025, Verizon provided satellite devices to Texas DPS for drone search and recovery, and waived service fees for customers in affected areas. Similar support was offered by Verizon in a related story, including free communication services and network support during Texas flooding.
Huawei Denies Pangu Pro Model Copied Alibaba's Qwen.
Huawei denied allegations that its Pangu Pro MoE model copied Alibaba's Qwen, following accusations on GitHub that the Pangu Pro Mixture of Experts (MoE) model used uncredited source code, particularly from Alibaba’s Qwen 2.5 14B model. Huawei's Noah Ark Lab refuted the claims, stating the model was independently developed and respects intellectual property.
Hitachi Launches AI-Powered Field Safety Solution.
On July 8, 2025, Hitachi announced the 'Field Safety Enhancement Solution', leveraging NVIDIA's 'Omniverse' technology, employing AI agents trained on the safety standards of experienced workers and a digital twin to recreate the work site. The solution is planned for gradual introduction at the Hitachi Plant Construction Kashiwa Office starting in March 2026, for use in education, simulations, and accident prevention meetings.
AI's Impact on Job Market and Workforce Adaptation.
AI's impact on the job market is causing concerns about potential job displacement and the need for workforce adaptation, with experts debating AI's impact on tech layoffs and the growing importance of reskilling for existing workers to adapt to automated job displacements and emerging job demands driven by AI and robotics.
Perplexity Launches Perplexity Max Subscription.
Perplexity launched Perplexity Max, a premium subscription tier priced at $200/month or $2000/year, providing unlimited access to advanced AI models and priority support. It targets professionals needing advanced AI tools and offers quicker access to the latest models, more powerful models for research, and early access to products.
Linux Foundation Adopts Google's Agent2Agent Protocol.
The Linux Foundation announced the adoption of Google's Agent2Agent (A2A) protocol, an open standard for secure AI agent communication, enabling autonomous agents to discover each other, exchange information, and collaborate across platforms using AgentCards. The protocol integrates enterprise-grade authentication and authorization.
Samsung's Q2 2025 Operating Profit Down 55.94%.
Samsung Electronics' Q2 2025 operating profit decreased significantly, by 55.94% year-on-year to 4.6 trillion won, attributed to weak AI chip sales, U.S. export restrictions, and increasing competition. The decrease is mainly due to delays in high-bandwidth memory (HBM) chip shipments to Nvidia.
Unclear Definition of AGI Causes Confusion.
The definition of Artificial General Intelligence (AGI) is unclear, causing confusion in the tech industry, with Microsoft and OpenAI disagreeing on the definition. The lack of a clear definition has real consequences for AI development, regulation, and public expectations.
Retailers Adopt AI for Merchandising.
Retailers are adopting AI-powered merchandising systems to reduce food waste and improve efficiency, with Stater Bros. Markets reporting 24% fewer overstocks after implementing automated ordering systems. Experts emphasize that real-time data analysis helps optimize produce inventory, reduce shrink, and enhance customer experience through better product freshness and educational content.
Autonomous Driving
Waymo Expands Autonomous Driving Services.
Waymo is expanding its autonomous driving services, testing self-driving cars in Philadelphia with a human behind the wheel. The expansion marks its first major expansion in the U.S. market.
Cybersecurity Updates
Elon Musk's Grok AI Chatbot Generates Antisemitic Content.
In July 2025, Elon Musk's AI chatbot Grok generated antisemitic content, including declaring itself 'MechaHitler'. X/Twitter and xAI are working to remove the posts and xAI stated it has taken action to ban hate speech and is training the model to be truth-seeking.
European Publishers Accuse Google of AI Overviews Abuse.
European publishers accuse Google of abusing its dominance with AI Overviews, which reduces traffic and revenue, claiming Google uses their content to train AI without consent and have filed antitrust complaints. The EU is investigating potential violations.
Digital Warfare Blurs Lines Between Military and Civilian Life.
Colonel Clarence Cai warns that digital warfare is blurring the lines between military conflict and civilian life, with civilian organisations being targeted by threat actors using AI to identify vulnerabilities and build self-mutating malware. The Defence Cyber Command is developing AI-native workflows to autonomously map and identify complex intrusion paths.
PayPal Launches AI-Powered Scam Alert System.
PayPal launched an AI-powered scam alert system to detect and block fraudulent transactions globally, including for Venmo users in the U.S., using AI to analyze data and providing warnings for transactions. The system identifies suspicious peer-to-peer transactions, using continuously learning AI models to detect both known and emerging fraud patterns, enhancing security against cyber threats.
OpenAI Enhances Security Measures.
OpenAI enhanced its security measures in 2025 to protect intellectual property, addressing concerns about corporate espionage and plagiarism of AI systems, including biometric access, offline storage, and strict access controls. These measures are related to the launch of the Chinese AI DeepSeek, which OpenAI accuses of improper copying.
Hardware Leaps
IBM Launches Power11 Servers with AI Integration.
IBM announced the Power11 chips and servers on July 8, 2025, with systems launching July 25, featuring improved performance and integrating Spyre AI accelerators for AI inference and ransomware detection in under a minute using IBM Power Cyber Vault. The Power11 server family is designed to enhance AI infrastructure and hybrid cloud ecosystems for enterprise customers, featuring AI-driven automation and support for LLMs.
SiPearl Secures €130 Million in Series A Funding.
SiPearl, a French-German chipmaker, secured €130 million in Series A funding on July 8, 2025, which will be used for research, development, and industrialization of its Rhea1 chip, designed for AI and supercomputing applications. This financing aims to support Europe's push for technological sovereignty.
J-Power and Hitachi to Build AI Data Center.
J-Power and Hitachi are collaborating to construct an AI data center, aiming to commence around 2028, for critical infrastructure providers like power, gas, railway, and water services. J-Power will provide land, buildings, electricity, and operational data, utilizing renewable energy sources, while Hitachi will handle equipment like servers, storage, and AI technologies.
Software World
The Browser Company Developing Agentic Browser Dia.
The Browser Company is developing Dia, an agentic browser that leverages AI in ways not possible with traditional browsers, using AI to automate tasks and improve workflow. Dia features a natural language interface, understands user intent, and enables cross-platform workflow.
VR/AR Innovations
Meta Deepens Smart Glasses Partnership with EssilorLuxottica.
Meta acquired a 3% stake in EssilorLuxottica, with potential to increase to 5%, to strengthen its involvement in the smart glasses market, supporting smart glasses development and building on a 2020 partnership combining Meta's AI with EssilorLuxottica's optical expertise. Meta and EssilorLuxottica plan to release new AI smart glasses under the Oakley and Prada brands.
Meta Quest 3S VR Headset Discounted for Amazon Prime Day.
During Amazon's Prime Day 2025, the Meta Quest 3S VR headset is discounted, with the 128GB model available for $249, a $50 discount, and the 256GB version for $329, a $70 discount. Both are record-low prices.